SMC-m-753 June .1970 em

THE SLJC LH2 TARGET CONTROL SYSTEM*

by

W. B. Pierce

Stanford Linear Accelerator Center Stanford University Stanford, California

Liquid Hydrogen (LH2) is utilized as a target for the high energy

electron beams, up to 20 GeV, available from the two-mile Stanford Linear

Accelerator because hydrogen is the purest form of protons attainable and

liquefaction increases the density of the protons available for bombardment.

This paper will deal with the necessary control systems required for the LH2

target, and special systems desired by experimenters.

Figure 1 is a sketch of the system. Components are situated at three

or four locations. The target assembly itself is located directly in the

beam line and is usually in an inaccessible high radiation area. The IX2

supply is located outside the high radiation area as close to the target as

possible and the controls are in an accessible location, near the experi-

menter's equipment.

The only component of the system of any interest to the experimenter is

the LH2 cell - target of the electron beam. Ideally the experimenter would

like a mass of LH2 suspended in a vacuum with a zero mass container. Since

this is not yet available, they have accepted- such materials as mylar

(between 5 to 15 mils thick depending upon the size and shape of target),

aluminum No. 6061 T-6 (3 to 6 mils), and stainless steel No. 314 (1 to 3

mils). The experimenter might be interested in any of the following .

parameters of this cell:

(Presented at the Third Internat ional Cmgenic Engineering Conference, iflay 25-27, 1970. Kongresshalle, West Berlin, Germany

* Work cupportcd by thcl IJnitcd States Atomic Energy Commission

1. LH2 level (full and empty)

2. Cell pressure Density determinants

3. Cell

An LH2 reservoir is situated within the vacuum chamber above the

LH2 cell. In the case of "direct-fill' targets a valve allows the

target to be gravity filled from the reservoir. In "condensation" type

targets the LH2 in the reservoir cools a heat exchanger which in turn

cools pressurized hydrogen gas in the cell which condenses. In this case

the LH2 cell is isolated from the reservoir.

The reservoir capacity is generally between 50 and 100 liters and

contains two independent techniques for measuring the LE2 level. One

technique uses carbon resistors to define a reservoir empty condition or

an overfill condition (> 90% full). This is basically the same circuitry

as that developed at Brookhaven 1 and utilizes a 300 R, l/2 watt Allen-

Bradley carbon resistor in a bridge circuit. At LH2 temperatures with no

current through the resistor, its resistance will increase to 'V 1100 Q.

In our circuit 150 milliwatts of power is dissipated in the resistor and

a comparator trip level is set to 'v 620 R. Greater resistance than 620 R

shows that the resistor is in liquid and less than 620 R shows gas. In

actual operation there is a discrete jump in resistance when the resistor

goes from 'a gaseous to a liquid ambient. This jump may be from 550 R to

700 R at the 150 mw dissipation level.

Figure 2 is a picture of the SIX! LH2 continuous level indicator.

This technique uses 32 silicon diodes connected in series and a 300 volt

power supply, with a series resistor to power the diode string. The

current in the string is monitored to indicate LH2 level. All preliminary

-2-

developmental work for this probe was done using LN2. Three types of

probes were considered for indicating LH2 level: series connected

diodes, series connected resistors, and a capacitance probe. Electronics

associated with the diodes or resistors is much simpler than the capaci-

tance probe which appeared to be most desirable from the mechanical

designer's viewpoint because there is no power dissipation and output is

insensitive to temperature changes, All three types were evaluated with

the following results:

1. The resistor probe: Af3 Repeatability was poor and the no was small in

comparison with the diode probe. The results of two test runs are shown

in Fig. 3. Due to the lack of repeatability of the probe we abandoned it

in favor of the diode probe.

2. The capacitance probe: Considerably more electronics are required:

an oscillator, a capacitance bridge, and detector circuitry. The factors

we considered in deciding against the use of a capacitor probe were:

The stray capacitance of feedthroughs, cables and connectors was

greater than the capacitance change due to the difference in dielectric

constant between the gaseous and liquid states. Since the dielectric

constant increases by 20% when the gas liquifies, the maximumcspsci-

tance change is only 20% of the base capacitance of the probe.

Extreme care must be taken to limit the capacitance to ground. These

problems were never completely solved. After considering the simplic-

ity of the diode probe we abandoned the capacitor probe. 2

3. The diode probe: Although we knew of no previous work done with

silicon diodes at IV LH2 temperatures, it appeared that they should exhibit

AR a large - AT

and therefore present a reasonably large amplitude signal for

processing. Initial tests were run with LN2. Thirty-two diodes were

-3-

serially connected and spaced along a perforated epoxy board approximately

25 cm long. The board was suspended vertically in a dewar with an indicating

float. LN2 WC%:6 added and both float levels and diode current

recorded. Figure 4 is a sketch of the test set up. The test results for

5 fill cycles using this circuit are shown in Fig. 5. The voltage supply

shown in Fig. 4 is adjusted to cause a current of 1 mA to flow through a

1 K R sensing resistor when the probe is totally immersed in I+. This

voltage averaged approximately 32 volts for 32 diodes. Allowing for a

higher voltage requirement when immersed in LH2, an automatic fill system

was designed for the target utilizing this diode probe, and the voltage was

set at 90 volts. When we first attempted to fill the reservoir with LH2,

the probe failed miserably. It oscillated between full on and full off

with lights flashing and relays chattering and the system looked like a

pinball machine. Diagnosing the problem, it became apparent that there

was insufficient voltage available from the power supply to sustain conduc-

tion at the LH2 temperature. All of our previous tests had been performed

at LN2 temperatures (- 77' K), but since LH2 (- 20' K) is so dangerous to

work with, we decided to perform some additional tests at Liquid Helium

(- 4' K) temperatures. A test set up consisting of a constant current

generator, series resistor, and a diode immersed in LHe was assembled.

The diode selected for use in our probe is type lN3C64 manufactured

by Computer Diode Corporation (No. CDC 6611). A voltage vs. current curve

was obtained at LHe temperatures (Fig. 6) and an examination of this curve

lead us to conclude that there appeared to be a negative resistance region

on the curve since the current increases as the voltage decreases. This

characteristic is similar to the V-I characteristic of gas tubes, which

-4-

require a certain voltage to ignite then regulate at a lower voltage.

This diode in LHe requires more than 12 volts before it will conduct but

requires less than 3 volts to sustain conduction currents of> 50 mu.

Next we decided to try the same test using LH2. The results are

shown in Fig. 7. The shapes of the curves are quite similar, the voltage

necessary to initiate conduction in LH2 is lower (approximately 9.5 volts).

Data points Ti:ere taken at a fixed current of

cryogenic liquid environments: LHe, IS2 and

1 mA for three

LN2 l

This curve is shown

in Fig. 8. From this curve a linear approximation over the range LHe

to LN2 can be made. The empirical relationship is:

V(lmA) = 12.2 - 0.15 T .

The basic diode equations 3 of I=1 do not hold in the 0

temperature range of LHe and LH2. Actually this is what appears to be

occurring at the diode junction:

(i) With no voltage impressed and no power dissipated within the diode, the

junction reaches thermal equilibrium at the temperature of the liquid cry-

ogen. After sufficient voltage is impressed on the diode to cause current

flow, power is dissipated and heat is generated at the junction. However,

as long as current flow is low enough the thermal conductivity of the .

diode removes the generated heat and the junction temperature remains low

. and fairly constant and the voltage required for conduction does not decrease.

As the current through the junction is increased though, power dissipation

increases until a point is reached where the thermal conductivity is not

enough to transfer heat from the junction fast enough, the junction tempera-

ture increases, and a lower voltage is required to sustain the same current

-5-

through the diode. Therefore, if the voltage is held constant, the current

will continue to increase - a negative resistance characteristic (see Fig. 7).

More effort is being planned to expand the available information on this

subject. One possible use for this phenomena is as a very simple but

extremely low power level sensor. If a voltage below the conductive threshold

is applied to the device when it is in a liquid cryogen, virtually no conduc-

tion occurs, and no power is dissipated. As the device comes out of the liquid

the heat transfer away from the junction is lessened and the junction temper-

ature rises, causing the device to conduct.

The Semiautomatic Vacuum System

A fully automated vacuum system for our LH2 target systems would not be

too difficult to design but the vacuum systems themselves have not reached

the state of the art where a fully automatic system would save time. It has

been demonstrated at SLAC that a vacuum system operates more efficiently if

an operator is included in the decision-making process and the electronics

is used to do three things:

Display the necessary information of vacuum levels and valve status. ' 1.

2. Prevent the operator from making an error by opening a valve when

it should not be opened.

3* Limiting cell pressures.

Figure 9 is a typical vacuum system used for the E-41 target. The logic restric-

tions placed on the valves are shown in the table on Fig. 9. Figure 10 is

the electronic circuitry utilized to perform this protective monitoring.

The circuit utilizes standard Motorola 5 volt integrated logic circuits

mounted on a plug-in printed circuit board. On initial installation there

were excessive unwanted multiple valve closings. To get the system to operate

-6- '

satisfactorily all DC voltages used to verify valve positions (clear

signals on Fig. 10) were bypassed with small 1.0 pfd capacitors located

right on the printed circuit board. The normal operating logic level for

these I.C's is 5 volts. A +25 volts supply on the chassis was used for

tripping the I.C. Is and was attenuated to a -t-5 volt signal on the printed

circuit board itself. Once these changes were made, the system functioned

beautifully.

The Air Flow Monitor

Each IX2 target at SLAC has an air blower associated with it capable

of exhausting the air surrounding the target within a minute. The early

targets had a two-speed blower which normally operated at low speed until

one of the hazardous atmosphere detectors sensed H2. The blowers were then

switched to the high speed to remove the H2 as quickly as possible. In

the latest target design the blowers operate at high speed continuously and

an independent airflow monitor circuit monitors this function. Should the

airflow rate drop below a preset level, alarms are actuated which call the

operator to the target. Figure 1lA is a picture of the airflow monitor;

Figures llB, C and D are schematics. The sensor itself consists of two

thermistors self-heated to a temperature of approximately 200°C, One

thermistor (the reference) is isolated from the airstream and the other is

directly in the flow path. The two thermistors are wired in a bridge circuit.

Airflow past the sensor thermistor cools it and offsets the bridge indicating

:a safe condition. 'During initial testing of this circuit, a cold mass

of air would cause the sensor thermistor to latch off with an extremely

long recovery. A special circuit to sensethis condition and boost the

- 7-

voltage available to the thermistor is incorporated in the control units.

Depending on the shroud covering the sensors, the device will sense a

very gentle airflow of a foot per second or, depending on the shroud, a

high velocity limited only by the ability of the device to sustain the

mechanical stresses.

H2 Detectors

The manufacturer selected to supply the hazardous atmosphere detectors

(W's) for the hydrogen target system was General Monitors, Inc. of

El Segundo,,California. These units operate satisfactorily as long as

they do not remain in a hydrogen contaminated atmosphere very long.

Continued operation in this environment results in deterioration of the

detecting element - a dangerous condition since hydrogen gas could be present

but undetectable. The detectors used on the hydrogen target are in a H2

environment only during a fault condition which should not last any longer

than a few minutes. Regular maintenance checks are made on the elements

to insure their proper operation and each target system has redundant

detectors. Detection of a hazardous atmosphere causes the automatic fill

system to latch off, audible alarms to sound and lights to flash locally

and at a remote control point.

Pressure Transducers

The standard pressure transducer we use is the Computer Instruments

Corporation (CIC) Model 4000. Its range is 0 to 60 psia with a 500 R

resistance element and is used with our vapor pressure bulbs for temperature

measurement of LH2 and rcmoting this information. The experimenters use

-8-

these devices to monitor the hydrogen cell temperature and, in addition,

when the temperature begins to increase, an alarm system alerts them to

a possible problem. Figure 12A is a picture of our temperature monitoring

circuits; Figures 12B, C and D are schematics.

System

Figure 13 is a picture of our first LH2 target control system. It

is the control system used for our less complicated direct fill targets:

the top chassis contains logic and switching for the vacuum system, status

summary for all the protection systems, and reservoir filling status and

control. The chassis immediately below the top one displays the LH2 level

status in both the reservoir and the cell. The diode probe is controlled

and read out from this chassis; so this chassis controls the LH2 reservoir

level automatically. Below this chassis is a recorder for monitoring signals

such as reservoir level. The two chassis immediately below the recorder are

the HAD controls. Below these two are the vacuum gauge controls. The

lowest chassis is an electrical circuit breaker for the rack and above this

is the air solenoid chassis with air solenoids to supply air pressure to

open and close valves mounted on'the target. These controls could normally

control a simple type "B" direct fill target. An experimenter wanting back-

ground information would occasionally like to direct the beam into the cell :

without any LH2 present. To do this, he closes valve V2 which closes off

the vent line from the cell and pressure increase due to heat leaks forces

the LH2 back into the reservoir. When he wishes LH2 to be present in the

cell he opens the valve and the cell will gravity fill from the reservoir.

Depending upon the size of the cell the emptying operation may take from

-9-

1 minute to an hour. The experimenters operating in our End Station A

have chosen to use a condensing type target with a separate dummy cell,

and, by moving the target vertically, they can select a dummy,or LH2

target with a minimum of delay (less than a minute). Once this system

was perfected they chose to have an ID2 target with a dummy cell along

with the LH2 system. In addition, we provided a total of 9 separate solid

targets for them. Our latest target installation (E-49a) has a total of

13 possible target selections with target selection control done in two

ways: (1) A n encoder is attached to a drive motor and reading correspond-

ing to certain target positions are determined. When a certain target is

desired, the corresponding number is dialed into the control chassis and

the target will be positioned automatically. (2) Selector switches,

corresponding to different possible targets, are depressed and the target

is moved until a second positioning switch is actuated.

After working with the control systems for a few years, we decided

to improve the system to make it easier and clarify the operation of

various valves. Figure 14 is the result of this attempt. The control

system displays and controls have been integrated into a flow diagram.

Figure 15 is included to show the complexity to which the control system

has grown.

- 10 -

BIBLIOGRAPHY

1. Liquefied Gas Level Indicator for H2 and N2, J. V. Cuttingham

and J. J. Kovarki.

Nuclear Instruments and Methods, Vol. 25 (1564), pp. 244-246.

2. United States Patent Number 3,465,587.

3* Transistor Engineering, A. B. Phillips, McGraw-Hill (1962).

4. Special Material Contributed by

John Mark Domingo Cheng Boris Golceff Randolph Champion Wesley Adams

.

2’ VAC PUNP D/SCHAf?GE

4” Gffa VENT TARCt? COVER

_--__-----------

14’ DUCT

I.25 Gff*

I I

NO NORt-fALL Y OPENED NC NORMALLY CLOSED P PRESSURE GAUCZ

~4 Pam SURE R~GUL A roR av RELIEF VALVE 60 BUR.5 T/NC D/SC PT PRESSURE

TRANSDUCER HAZARDOUS

HAD A TMOSPffERE DfrECTORS

FIG.

1598A7

.--IX2 target system.

EMPTY 25

FULL 0

0 Fill #I x Fill #2 A Fill #3 •I Fill #4 x -

4 -

I I I I I I I I I I I I I I ri I

0 I 23456789 OUTPUT VOLTAGE - 159881

FIG. 3--Resistor probe voltage vs. L.N2 level. (47 each,109 , +W resistors in LIT*.)

ADJUSTABLE POWER

SUPPLY

- + = 31 VOLTS I I

Ocm

CM SCALE

25cm

-FLOAT

I59886

FIG. b--Diode test setup.

25

20

15

IO

5

FULL 0

0

X

A

0

Fill #I Fill #2 Fill #3 Fill #4

0 3 5 6 8 9

1

DIODE CURRENT (mA) - 159882

FIG. 5--Diode probe current vs. LI$ level. (32 each, CM: 6611, four separate fills.)

g lo-3 = E K s

lo-4 -

lo+ 7

I 1 I I I 1 I -

CDC 6611 Si DIODE z LH, ENVIRONMENT Z

CURRENT I xlo-6

9 x 10-6 I~xIO-~

1ooxlo-6 160 x IO+

500 x lo-6 I xIo-3

1.5 x 10-3 2.0x lo-3

5x lo-3 8x lO-3

IOX lo-3 12x lo-3

13.3x lo-3 20x 10-3 50x 10-3

IOOX lo-3

VOLTS

12.77 12.6 12.6 12.3 - 12.4 12.3 - 12.4 I I.9 - 12.0 ll.6- II.8 ll.5- II.6 II.1 - II.2 9.9 -10. I 8.8 - 9. I 8.2- 8.3 7.7- 7.9 6.6 - 8.0 5.4 - 6.7 2.7 - 2.8 2.4 - 2.5

lo+ I I I I 1 I I I I I 2 4 6 8

VOLTAGE IO

159883

FIG. 6 --Yoltage vs. current. (CDC 6611, Si diode, LHe environment.)

d . tn

P a l-J*

H

.

P 0

CD

C 6

611

DIO

DE

,CU

RR

ENT

(AM

PS)

14 13 12 I I

IO 9 8 7 6 5 4 3 2 I

0

I I I I I I I I I I I I I I I I I I

4.3O LH, 11.7 20° LH2 9.0 76O LN7 1.3

-

ar Approx. V=l2.2- kT

k ‘” 10~7-4.7 h ,r- .-.. . -.

Z-----L - =Ol9 I AT 50-10 W.IW I

\ \ \ V=l2.2-O.l5T

-

LD 3 - AL

L

\ I \ LO \ 2

Lb r\ - - LH e LH 2 LN e

- I I I I I I I I I I Ii I I I I

0 IO 20 30 40 50 60 70 80 90 IO0 TEMPERATURE “K

FIG. 8--Volts required to initiate 1 mA conduction in a CDC 6611 silicon diode VS. cryogenic temperatures of liquid cryogen.

.:

TO QOOF VENT

MP- I

FESEQVOIff

SCATTEelN($ CHAMBEQ

5. Opn V-1,. (D.P. PDreLily Vacuum)

6. Open V-10 (D.P. 1nolltic.i)

*. V.C"um It PIG6 above *et point. B. Thei-mo-s"itch O.k. c. HP-1 0".

A. MP-1 running. B. V-3 clmcd. c. v-11 Cllmed.

A. w-1 running. B. V-4 closed. c. v-11 closed. D. v-10 closed.

A. MP-1 running. B. V-k closed. c. v-3 closed.

A. V.C”um at FTC-6 below set pin!. 8. v-3 closed. c. DP-1 running.

A. KP-1 Ofl. 8. v-3 closed. c. v-4 closed. D. v-ll clmcd.

1598A9

FIG. +-Typical vacuum system.

L----J

L-----J

1598A10

FIG. lo--Integrated circuit vacuum logic P.C. board.

m

0. n

.

INTERLOCk d

ILUMIN~TION

2w 3>*

I 4 >Fl-+ 5>)+-- w l-

I 1

f

NOTFS:

izi TEST POINTS TPI + TP2 MOONTED ON FX. FmVtD 1598Al2

FIG. lib--Hz vent air flow monitor.

TPI 1

P- +&

RIO

QL 2”L90! P-

111‘ I.,K

1598A13

FIG. llc--Comparator driver P.C. board - H2 vent air flow monitor.

S A1ddt-G

2l3M

Od

c--.._-_---- -

---__I--- --_

I ---

> A-

I* 1

- I

Q

I ON

LlldLnO

03~vln93a

I h-lddns a3M

Od

437p)o Ow

Ol~i

_-__-____ --------

---5 -

- -.-...A

.9

.

1598A16

FIG. 12b--Cell temperature indicator.

m

zi4 U

O_m

--k

A c

(I-

-B-

lh

1598A19

FIG. 13--Typical "B" target control.

FIG. lb-latest model target control. 1398AZO